Understanding and Eliminating the Detrimental Effect of Endogenous Thiamine

bioRxiv preprint doi: https://doi.org/10.1101/753525; this version posted September 4, 2019. The copyright holder for this preprint (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made available under aCC-BY-NC-ND 4.0 International license.

1 Understanding and eliminating the detrimental effect of endogenous thiamine

2 auxotrophy on metabolism of the oleaginous yeast Yarrowia lipolytica

4 Caleb Walker,1,§ Seunghyun Ryu,1,§ Richard J. Giannone,2 Sergio Garcia,1 Cong T. Trinh1,*

5 1Department of Chemical and Biomolecular Engineering, The University of Tennessee Knoxville,

6 TN 37996

7 2Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, TN 37831

9 §Equal contributions

10 *Corresponding author

11 Email: [email protected]

12 Tel: 865-974-8121

16 ABSTRACT

17 Thiamine is an essential vitamin that functions as a cofactor for key enzymes in carbon and energy

18 metabolism for all living cells. While most plants, fungi and bacteria can synthesize thiamine de

19 novo, the oleaginous yeast, Yarrowia lipolytica, cannot. In this study, we used proteomics

20 together with physiological characterization to understand key metabolic processes influenced

21 and regulated by thiamine availability and identified the genetic basis of thiamine auxotrophy in

22 Y. lipolytica. Specifically, we found thiamine depletion results in decreased protein abundance

23 of the lipid biosynthesis pathways and energy metabolism (i.e., ATP synthase), attributing to the

24 negligible growth and poor sugar assimilation observed in our study. Using comparative

25 genomics, we identified the missing gene scTHI13, encoding the 4-amino-5-hydroxymethyl-2-

26 methylpyrimidine phosphate synthase for the de novo thiamine synthesis in Y. lipolytica, and

27 discovered an exceptional promoter, P3, that exhibits strong activation or tight repression by low

28 and high thiamine concentrations, respectively. Capitalizing on the strength of our thiamine-

29 regulated promoter (P3) to express the missing gene, we engineered the first thiamine-

30 prototrophic Y. lipolytica reported to date. By comparing this engineered strain to the wildtype,

31 we unveiled the tight relationship linking thiamine availability to lipid biosynthesis and

32 demonstrated enhanced lipid production with thiamine supplementation in the engineered

33 thiamine-prototrophic Y. lipolytica.

35 IMPORTANCE

36 Thiamine plays a crucial role as an essential cofactor for enzymes in carbon and energy

37 metabolism of all living cells. Thiamine deficiency has detrimental consequences on cellular

38 health. Yarrowia lipolytica, a non-conventional oleaginous yeast with broad biotechnological

39 applications, is a native thiamine auxotroph, whose effects on cellular metabolism are not well

40 understood. Therefore, Y. lipolytica is an ideal eukaryotic host to study thiamine metabolism,

41 especially as mammalian cells are also thiamine-auxotrophic and thiamine deficiency is

42 implicated in several human diseases. This study elucidates the fundamentals of thiamine

43 deficiency on cellular metabolism of Y. lipolytica and identifies genes and novel thiamine-

44 regulated elements that eliminate thiamine auxotrophy in Y. lipolytica. Furthermore, discovery

45 of thiamine-regulated elements enables development of thiamine biosensors with useful

46 applications in synthetic biology and metabolic engineering.

48 Keywords: Yarrowia lipolytica; thiamine metabolism; thiamine auxotroph; thiamine prototroph;

49 thiamine-regulated promoters; lipid production; thiamine deficiency

53 INTRODUCTION

54 Thiamine, or vitamin B1, was the first B vitamin discovered. Its activated form, thiamine

55 pyrophosphate (TPP), functions as a cofactor for key enzymes in carbon metabolism including

56 glycolysis, citrate cycle, pentose phosphate and branched chain amino acid pathways (Figure 1)

57 (1). TPP-dependent enzymes, including pyruvate dehydrogenase (PDH), α-ketoglutarate

58 dehydrogenase (KGDH), transketolase (TKL), branched chain α-ketoacid dehydrogenase (BCKDC)

59 and acetolactate synthase (AHAS), are essential for maintaining cell growth and preventing

60 metabolic stress (2, 3). Specifically, PDH links the glycolysis and citrate (Krebs) cycle by catalyzing

61 the conversion of pyruvate into acetyl-CoA (4)(5). KGDH, the rate-limiting step of the citrate

62 cycle, converts α-ketoglutarate to succinyl-CoA (6, 7). TKL is responsible for the last step of the

63 pentose phosphate pathway where it interconverts pentose sugars to glycolysis intermediates

64 (8). The pentose phosphate pathway is critical for production of ribose (i.e., RNA), precursor

65 metabolites for aromatic amino acid pathways, and reducing equivalents (i.e., NADPH) necessary

66 for maintaining redox balance and lipid synthesis. Hence, TKL activity is critical for RNA, protein,

67 and lipid production while preventing oxidative stress (9, 10). Meanwhile, AHAS and BCKDC are

68 responsible for the synthesis and degradation of branched chain amino acids (i.e., valine, leucine

69 and isoleucine), respectively (11)(12).

70 In mammals, thiamine deficiency affects the cardiovascular and nervous systems resulting

71 in tremors, muscle weakness, paralysis and even death (13)(14). Thiamine deficiency can occur

72 from inadequate intake, increased requirement, or impaired absorption of thiamine (15).

73 Biochemical consequences of thiamine deficiency result in failure to produce adenosine

74 triphosphate (ATP), increased acid production (i.e., lactic acid), decreased production of acetyl-

75 compounds (i.e., acetylcholine), neurotransmitters (i.e., glutamate, aspartate, aminobutyric

76 acid), reduced nicotinamide adenosine dinucleotide phosphate (NADH), defective ribonucleic

77 acid (RNA) ribose synthesis, and failure to break down branched chain carboxylic acids (i.e.,

78 leucine, valine, isoleucine)(16-18).

79 While mammals require nutritional supplementation of thiamine from dietary sources,

80 most bacteria, fungi and plants can synthesize thiamine endogenously. One of these exceptions

81 is the thiamine-auxotrophic oleaginous yeast, Yarrowia lipolytica, which has recently emerged as

82 an important industrial microbe with broad biotechnological applications due to its GRAS status

83 (19), metabolic capability (20-24), and robustness (25-27). Hence, Y. lipolytica is an ideal

84 eukaryotic host to study the fundamental effects of thiamine deficiency on cellular health.

85 Currently, it is not well understood why thiamine deficiency causes failure of thiamine

86 biosynthesis in Y. lipolytica and how this deficiency might directly affect other processes (e.g.,

87 lipid biosynthesis, energy metabolism, etc.) beyond what is already known.

88 Interestingly, Y. lipolytica’s auxotrophy has been exploited for enhanced production of

89 organic acids (i.e., pyruvate and KGA) by reducing activities of PDH and KGDH under thiamine-

90 limited conditions (28, 29). Consequently, cell growth is negatively affected by thiamine

91 limitation and completely prevented when thiamine is depleted from the media. Not

92 surprisingly, the genes in thiamine metabolism are tightly regulated by thiamine concentrations

93 (30, 31). To this end, numerous thiamine-regulated promoters have been discovered in yeast

94 that enable control of genetic expression by adjusting thiamine concentration in the culture

95 media (32-34). However, endogenous thiamine-regulated promoters have not yet been

96 discovered in Y. lipolytica. Furthermore, fundamental understanding of how thiamine deficiency

97 inhibits metabolism and hence growth of Y. lipolytica is still lacking.

98 In this study, we shed light on the effect of thiamine deficiency on cellular metabolism in

99 the thiamine auxotroph Y. lipolytica. We identified the missing gene, scTHI13, in the de novo

100 thiamine biosynthesis pathway of Y. lipolytica and discovered a thiamine-regulated promoter,

101 P3, that increases expression in low thiamine concentrations. By employing P3 to control the

102 expression of scTHI13, we engineered thiamine prototrophy for the first time in Y. lipolytica. This

103 enabled us to elucidate the relationship between thiamine availability and ATP and lipid

104 biosynthesis and revealed enhanced lipid production in the engineered thiamine-prototrophic Y.

105 lipolytica.

106

107 RESULTS

108 The effects of thiamine deficiency in Y. lipolytica

109 Cell growth and organic acid production are influenced by thiamine limitation and

110 depletion. To demonstrate the effects of thiamine limitation, we characterized the thiamine-

111 auxotrophic Y. lipolytica (YlSR001) in 0, 0.5 and 400 µg/L thiamine (Figure 2A). As expected, cell

112 growth was inhibited by limited (0.5 ug/L thiamine) and depleted (0 ug/L thiamine)

113 concentrations of thiamine but restored in high thiamine (400 ug/L thiamine) containing media

114 (Figure 2B). Glucose consumption profiles were closely coordinated with cell growth (Figure 2C).

115 Within the first 24 hr, only ~2 g/L glucose was assimilated without thiamine while thiamine-

116 limited cultures consumed ~3 g/L glucose (Figure 2C). After 24 h, growth and glucose

117 consumption were completely stalled under the no thiamine condition. In contrast, cells

118 continued to utilize glucose even though growth was significantly inhibited.

119 Next, we characterized pyruvate and KGA production since the enzymes converting these

120 organic acids, PDH and KGDH respectively, are TPP-dependent. In high thiamine media, Y.

121 lipolytica demonstrated slight KGA production (~0.5 g/L KGA) and no pyruvate accumulation

122 (Figure 2D, 2E). Not surprisingly, we observed substantial pyruvate titers in thiamine limited (~4

123 g/L pyruvate) and depleted (~2 g/L pyruvate) media (Figure 2E) as PDH is unable to efficiently

124 convert pyruvate to acetyl-CoA. Similarly, KGA accumulation was enhanced in thiamine-limited

125 media (2 g/L KGA) but KGA titers in thiamine-depleted media were comparable to high thiamine

126 media likely due to decreased flux from glycolysis to the TCA cycle through deficient acetyl-CoA

127 production via PDH (Figure 2D). Taken together, Y. lipolytica requires thiamine supplementation

128 for cell growth and carbon assimilation but produces organic acids under thiamine-limited

129 conditions.

130 Proteome of thiamine depletion reveals perturbation of critical metabolic pathways

131 related to cellular growth. Next, we investigated the proteome of the thiamine-auxotrophic Y.

132 lipolytica growing in 0 and 400 µg/L thiamine (Figure 3A). Across 2 exponential time points, we

133 identified 535 upregulated and 515 downregulated proteins (i.e., log2 fold change > |1|) in

134 response to thiamine deficiency (Figure 3B). First, we looked at metabolic enzymes that require

135 TPP as a cofactor including PDH, KGDH, transketolase (TKL), branched chain α-ketoacid

136 dehydrogenase (BCKDC) and acetolactate synthase (AHAS) (Supplementary Table S1).

137 Interestingly, all proteins encoding subunits (E1-E3) of BCKDC were upregulated in thiamine

138 depletion (Supplementary Table S1). However, none of the other TPP-requiring proteins were

139 upregulated in media lacking thiamine except for dihydrolipoamide dehydrogenase (E3), which

140 serves as a subunit for BCKDC, PDH and KGDH (Supplementary Table S1).

141 We then mapped the 535 upregulated- and 515 downregulated proteins to their

142 respective KEGG pathways (Figure 3C). Thiamine deficiency resulted in downregulation of

143 proteins involved in nucleic acid metabolism (i.e., pyrimidine and purine metabolism) and genetic

144 processes (i.e., ribosome and DNA replication). Without thiamine, cells also exhibited decreased

145 protein abundance in lipid metabolism which includes the synthesis of terpenoids, steroids and

146 glycerophospholipids. In contrast, thiamine deficiency resulted in upregulation of proteins

147 contained in carbohydrate (i.e., glycolysis and citrate cycle), tetrapyrrole, biotin, aromatic (i.e.,

148 tryptophan, phenylalanine), branched chain (i.e., leucine, isoleucine, valine), and other amino

149 acid (i.e., glycine, serine, threonine) metabolic pathways. Thiamine deficiency also affected

150 proteins associated with the electron transport chain (ETC), amino acid (i.e., arginine, proline,

151 glutamate, glutamine, methionine, cysteine) biosynthesis and thiamine metabolic pathways.

152 A closer look at proteins involved in the ETC and thiamine metabolism revealed

153 interesting features. First, thiamine-deficient cells exhibited increased protein abundances for all

154 four complexes of the ETC but decreased protein abundance for ATP synthase that is important

155 for ATP generation (Figure 3D). The phenomenon correlates with the stalled assimilation of

156 glucose that is ATP-dependent when cells grow in the absence of thiamine. Second, all but one

157 of the proteins in thiamine metabolism were differentially regulated between thiamine-sufficient

158 and thiamine-depleted cultures (Figure 3E). Without thiamine, we observed increased

159 abundance of proteins in the upper branch of thiamine synthesis but decreased abundance in

160 proteins converting thiamine monophosphate into thiamine and TPP into thiamine triphosphate.

161 The only detected protein in thiamine metabolism that was not differentially translated was

162 thiamine kinase (Thi90p), which is responsible for the conversion of thiamine into its activated

163 diphosphate form, TPP. Taken together, thiamine concentration influences regulation of

164 thiamine metabolism, whereby thiamine deficiency severely affects central carbon metabolism

165 and elicits increased protein abundance for most carbohydrate, amino acid and energy pathways

166 but decreased protein abundance for lipid metabolism, nucleotide metabolism and specifically,

167 ATP synthase.

168 Restoring thiamine prototrophy in Y. lipolytica

169 Thiamine metabolism in Y. lipolytica is incomplete. Next, we investigated the native

170 thiamine metabolism of Y. lipolytica to elucidate underlying genetic deficiency causing thiamine-

171 auxotrophic behavior. We compared the thiamine biosynthesis pathways between the well-

172 characterized thiamine-prototrophic Saccharomyces cerevisiae and our thiamine-auxotrophic Y.

173 lipolytica. Between the two organisms, only one missing gene was identified (scTHI13), encoding

174 4-amino-5-hydroxymethyl-2-methylpyrimidine phosphate synthase that converts histidine and

175 pyrodixine into hydroxymethylpyrimidine pyrophosphate (HMP-PP), which is likely required for

176 the de novo thiamine synthesis in Y. lipolytica (Figure 3E).

177 Constitutive expression of the missing thiamine gene does not effectively restore

178 prototrophy. To enable the de novo biosynthesis of thiamine in Y. lipolytica, we constructed a

179 vector to express scTHI13 under the constitutive promoter, TEF, frequently used for genetic

180 overexpression in Y. lipolytica. Unexpectedly, expression of scTHI13 with the TEF promoter only

181 restored the de novo TPP biosynthesis after days of adaptation in thiamine-depleted media

182 (Supplementary Figure S1). Additionally, cell growth was highly deviated between replicates for

183 both times when this experiment was conducted (Supplementary Figure S1). Though the results

184 were somewhat promising, we endeavored to engineer a true thiamine-prototrophic Y. lipolytica.

185 Bioinformatics of thiamine-responsive genes reveals highly regulated thiamine

186 promoter. We hypothesized that expression of scTHI13 was weak since the constitutive TEF

187 promoter (35, 36) is dependent on cell growth (37, 38) and thiamine deficiency prevents cell

188 growth. To this end, we aimed to find a promoter responsive to thiamine deficiency. Through

189 BlastP and orthology analysis using the well-studied thiamine-regulated genes from Pichia

190 pastoris, Saccharomyces cerevisiae and Schizosaccharomyces pombe, we identified three

191 candidate genes that are putatively regulated by thiamine in Y. lipolytica (Supplementary Table

192 S2). The P1 gene (YALI0E04224g) encodes a putative thiaminase that might exhibit

193 hydroxymethylpyrimidine kinase/phosphomethylpyrimidine kinase activity for conversion of

194 thiamine into 4-amino-5-hydroxymethyl-2-methylpyrimidine diphosphate. The P3 gene

195 (YALI0A09768g) encodes a putative cysteine-dependent adenosine diphosphate thiazole

196 synthase that is involved in the biosynthesis of thiazole, a thiamine precursor. Thiazole synthase

197 converts NAD+ and glycine into ADP-5-ethyl-4-methylthiazole-2-carboxylate, a thiazole

198 intermediate. Though the function of P2 (YALI0C14652g) has yet to be characterized, it harbors

199 a NMT1 domain (Pfam:PF09084) which is required for biosynthesis of the pyrimidine moiety of

200 thiamine (39); hence, P2 might be thiamine-regulated.

201 To determine if these promoters (i.e., P1, P2, and P3) are thiamine-regulated, real-time

202 PCR was conducted to quantify mRNA levels of P1, P2, and P3 genes from YlSR001 grown in low

203 (0.5 µg/L) and high (500 µg/L) concentrations of thiamine (Figure 4A). The expression levels of P1

204 and P3 genes were activated in low thiamine but inactivated in high thiamine. Remarkably, P3

205 exhibited substantially increased expression relative to P1 by 7.60 ± 1.51-fold in low thiamine.

206 We did not observe transcriptional expression of P2 in low or high thiamine concentrations.

207 We next investigated plasmid expression of these 3 putative thiamine-regulated

208 promoters from the Y. lipolytica strains that harbor a humanized renilla green fluorescent protein

209 (hrGFP)-expressing gene under the control of either P1, P2, or P3 promoter (strains YlSR1002,

210 YlSR1003 and YlSR1004, respectively). For controls, we also built the native, constitutive TEF

211 promoter and the heterologous NMT1 promoter from S. pombe that has been well-studied and

212 applied as a thiamine-regulated promoter (40). The hrGFP intensity was monitored from cells

213 grown at low (1 µg/L) and high (10 mg/L) concentrations of thiamine (Figure 4B). As expected,

214 both P1 and P3 promoters were activated in low thiamine and inactivated in high thiamine.

215 Under low thiamine conditions, the hrGFP intensity under promoter P3 was exceptionally higher

216 than P1 and TEF promoters by 10.04 ± 0.46-fold and 2.82 ± 0.13-fold, respectively. Not

217 surprisingly, the activity of TEF promoter was much greater in high thiamine than in low thiamine.

218 The NMT1 promoter, however, showed no activity in low thiamine and limited activity in high

219 thiamine.

220 We further investigated the sensitivity of the P3 promoter over a range of low thiamine

221 concentrations (0.5-50 µg/L) (Figure 4C). Encouragingly, the P3 promoter exhibited tight

222 regulation across all incremental thiamine concentrations. Of note, the activity of the P3

223 promoter was not affected by cell growth (Figure 4D). Taken together, these data indicate that

224 the endogenous thiamine-regulated promoters, P1 and P3, can be applied to strain engineering

225 efforts and are responsive to low thiamine concentrations.

226 Thiamine prototrophy is restored with thiamine regulated promoter. To restore the de

227 novo thiamine synthesis in Y. lipolytica, we constructed the vector pSR075 to express the missing

228 thiamine gene, scTHI13, with the thiamine-responsive promoter, P3 (i.e., YlSR1006). Remarkably,

229 this construct demonstrated robust and reproducible growth in media lacking thiamine (Figure

230 5). Cell growth and glucose uptake were similar despite with low to no added thiamine (Figure

231 5A, 5B), but organic acid production was affected. Organic acid production in low (0.5 µg/L) and

232 depleted (0 µg/L) thiamine conditions was diminished during the stationary phase (Figure 5C,

233 5D). Meanwhile, KGA accumulation only manifested in high (400 µg/L) thiamine (Figure 5C).

234 Taken together, the thiamine prototroph strain created here, YlSR1006, grows reproducibly

235 irrespective of thiamine concentrations but produces no organic acids in thiamine-limited media.

236 Lipid accumulation is influenced by thiamine concentrations. Finally, we investigated

237 the relationship between thiamine availability and lipid accumulation. We cultured both

238 thiamine-auxotrophic (wildtype) and our engineered thiamine-prototrophic (YlSR1006) strain

239 with 0 g/L and 400 g/L thiamine in MpA (no nitrogen limitation, Figure 6A) and lipid production

240 (nitrogen limitation with C:N = 100, Figure 6B) media. In non-nitrogen limited media, YlSR1006

241 produced more lipid than the wildtype grown with 400 g/L thiamine supplementation (Figure

242 6A). Interestingly, YlSR1006 grown without thiamine accumulated lipid similar to the wildtype

243 with 400 g/L thiamine. However, under nitrogen-limitation, both the wildtype and YlSR1006

244 strains showed similar lipid production profiles when 400 g/L thiamine was supplemented in

245 growth medium (Figure 6B). In thiamine-lacking medium, while no lipid accumulation was

246 expected for the wildtype, YlSR1006 was able to accumulate 3.68 ± 0.22 lipid %DCW, which was

247 ~50% less than YlSR1006 supplemented with 400 µg/L thiamine. Taken together, thiamine

248 supplementation increased lipid production even for thiamine-prototroph YlSR1006, indicating

249 thiamine plays a critical role for lipid biosynthesis in Y. lipolytica.

250 DISCUSSION

251 In this study, we observed the effects of thiamine deficiency on growth, sugar

252 consumption, organic acid production, and proteome of the thiamine-auxotrophic Y. lipolytica.

253 The activated form of thiamine, TPP, is an important cofactor for enzymes involved in vital cellular

254 functions including energy metabolism (41), reducing oxidative and osmotic stresses (42), and

255 catabolism of sugars (43). Hence, the consequences of thiamine deficiency are caused by the

256 reduced activity of TPP-dependent enzymes (i.e., PDH, KGDH, TKL, AHAS and BCKDC), which

257 ultimately leads to cell death. A comprehensive model that depicts the detrimental effect of

258 thiamine deficiency on metabolism leading to cell death in Y. lipolytica is summarized in Figure

259 7.

260 Loss of PDH and KGDH activities result in poor growth, limited carbon assimilation and

261 accumulation of pyruvate and KGA (Figure 2). Failure to metabolize pyruvate via PDH also inhibits

262 production of acetyl-CoA, the precursor metabolite of the TCA cycle. The TCA cycle is further

263 inhibited by reduced KGDH activity, preventing synthesis of NADH required for oxidative

264 phosphorylation (i.e., ETC). Notably, in the model yeast, Saccharomyces cerevisiae, deletion of

265 KGDH is known to prevent respiratory growth (44), but the exact mechanism is not well

266 established. In our study, proteomic analysis shows that thiamine-deficient cells increased

267 protein abundance for the first four complexes of oxidative phosphorylation (i.e., NADH

268 dehydrogenase, succinate dehydrogenase, and cytochrome c reductase/oxidase) but decreased

269 protein abundance for ATP synthase (Figure 3D).These new findings indicate that thiamine

270 deficiency negatively affects respiratory energy metabolism caused by a malfunctioning TCA

271 cycle via inhibition of PDH and KGDH as observed by inhibited growth and reduced sugar uptake

272 in thiamine-depleted cells.

273 Y. lipolytica also decreased protein abundances in glycerophospholipid, terpenoid

274 backbone and sterol biosynthesis pathways in thiamine-deficient cells (Figure 3C). These

275 phenomena are likely consequences of reduced PDH activity (i.e., reduced pools of acetyl-CoA)

276 and likely affected by NADPH production by the pentose phosphate pathway. In the pentose

277 phosphate pathway, TKL interconverts pentose sugars and hexose sugars, the later which serve

278 as glycolytic intermediates (e.g., fructose-6P, glyceraldehyde-3P). Hence, loss of TKL activity

279 likely effects the production of ribose-5P and NADPH which are required for synthesis of lipids,

280 RNA, DNA, purines, pyrimidines, and antioxidants (45). Interestingly, loss of TKL activity also

281 prevents production of erythrose-4P, impeding synthesis of folate and aromatic amino acids (i.e.,

282 phenylalanine, tyrosine, and tryptophan) as previously observed in TKL-deletion mutants of S.

283 cerevisiae (46).

284 In our study, Y. lipolytica also responded to thiamine deficiency by increasing protein

285 abundance of enzymes involved in branched chain α-amino acid metabolism (i.e., valine, leucine

286 and isoleucine). Consistent with the previous studies with S. cerevisiae, deletion of BCKDC results

287 in branched chain α-amino acid auxotrophic phenotypes (47). Interestingly, BCKDC was the only

288 TPP-requiring enzyme with all 3 subunits upregulated in our thiamine-depleted Y. lipolytica cells.

289 However, this finding might be the result of leucine supplementation in the media since our Y.

290 lipolytica strain is leucine auxotrophic. Taken together, thiamine deficiency inhibited cell growth

291 severely limiting energy production, amino acids (i.e., aromatic and branched chain) and lipid

292 synthesis, ultimately leading to cell death (Figure 7).

293 Despite thiamine-mediated growth inhibition, Y. lipolytica exhibited strong upregulation

294 of proteins contained within thiamine metabolism in response to thiamine depletion (Figure 3E).

295 Interestingly, one of the enzymes upregulated in thiamine depletion, cysteine-dependent

296 adenosine diphosphate thiazole synthase (YALI0A09768g), is promoted by our thiamine-

297 regulated promoter, P3. While various constitutive and inducible promoters are available for Y.

298 lipolytica (48-51), our tightly regulated, thiamine-responsive P3 promoter is especially useful for

299 strong, inducible expression in low thiamine conditions. Under optimized conditions, the activity

300 of pP3 was 2.82 ± 0.13-fold higher than TEF (Figure 4B). Hence, gene overexpression using the

301 P3 promoter is highly desirable over the constitutive TEF promoter in low thiamine

302 concentrations. This was demonstrated by restoring thiamine prototrophy in Y. lipolytica using

303 P3 promoter (Figure 5), while the TEF promoter was not strong enough to accomplish stable

304 thiamine prototrophy (Supplementary Figure S1).

305 Surprisingly, thiamine prototroph was restored by overexpressing 1 single gene, absent

306 from the native metabolism of Y. lipolytica, for the de novo synthesis of thiamine (Figure 3E,

307 Supplementary Figure S1). This begs the question, why does Y. lipolytica lack this gene? We

308 hypothesize that most Y. lipolytica strains are isolated from thiamine-rich sources (e.g., sausage,

309 oats, plants) (52) while most popular yeasts, including S. cerevisiae, are isolated from sugar-rich

310 sources (e.g., fruits, molasses, sugarcane) (52). Remarkably, we observed enhanced lipid

311 production by thiamine supplementation in both thiamine-auxotrophic and prototrophic strains,

312 suggesting a relationship between lipid production and thiamine availability (Figure 6A, 6B). This

313 relationship shows promise for industrial applications of neutral lipid production and establishes

314 a novel direction for increasing lipid production in Y. lipolytica.

315

316 MATERIALS AND METHODS

317 Plasmids and strains

318 The list of plasmids and strains used in this study is available in Supplementary Table S3.

319 Plasmid pSR005 carrying a humanized renilla green fluorescent protein (hrGFP) was constructed

320 by Gibson assembly method (53) with hrGFP and pSL16-CEN1-1-227 (54). The hrGFP gene was

321 amplified using the primers hrGFP_Fwd and hrGFP_Rev from pBABE GFP (Addgene plasmid

322 #10668). The backbone pSL16-CEN1-1-227 was amplified with the primers pSL16_Fwd and

323 pSL16_Rev.

324 Next, various promoters, TEF(404) (50), NMT1 (55), P1(1000), P2(1000), and P3(1000)

325 were inserted into pSR005 by Gibson assembly method. Each TEF(404), P1(1000), P2(1000), and

326 P3(1000) promoter region (size of each promoter is in parenthesis) was amplified using the

327 primers PTEF_Fwd/PTEF_Rev, PP1_Fwd/PP1_Rev, PP2_Fwd/PP2_Rev, and PP3_Fwd/PP3_Rev,

328 respectively from the genomic DNA of Y. lipolytica ATCC MYA-2613. The NMT1 promoter region

329 was amplified using the PNMT1_Fwd/PNMT1_Rev primers from the genomic DNA of S. pombe (kindly

330 provided by Dr. Paul Dalhaimer, Department of Chemical and Biomolecular Engineering,

331 University of Tennessee Knoxville, TN, USA). The backbone pSR005 was amplified by using the

332 primers pSR005_Fwd/pSR005_Rev. The constructed plasmids are pAT32y, pSR068, pSR071,

333 pSR072, and pSR073 (Supplementary Table S3).

334 The plasmid pSR074 was constructed by assembly of (i) THI13sce, amplified using the

1 335 primers THI13sce_Fwd /THI13sce_Rev from the genomic DNA of S. cerevisiae and (ii) pSR008

336 backbone, amplified using the primers pSR008_Fwd/pSR008_Rev. The plasmid pSR075 was

337 constructed by replacing the hrGFP gene with the THI13sce gene from pSR073. The THI13sce

2 338 gene was amplified by using the primers THI13sce_Fwd /THI13sce_Rev from the genomic DNA of

339 S. cerevisiae and assembled with the PP3 promoter carrying backbone, amplified by using the

340 primers pSR008_Fwd/pSR073_Rev.

341 The Y. lipolytica ATCC MYA-2613, obtained from ATCC strain collection, was used as a

342 parent strain. YlSR101, YlSR109, YlSR1001 and-YlSR1006 (Supplementary Table S3) strains were

343 generated by transforming the corresponding plasmids via electroporation (56). Each plasmid

344 was transferred into Y. lipolytica YlSR001 via electroporation to generate strains used in this

345 study. The YlSR109, YlSR1001-YlSR1004 strains were confirmed by the respective promoter

346 binding forward primer along with hrGFP_Rev. The YlSR1005 and YlSR1006 strains were

347 confirmed by TEF(-100)_Fwd or P3(-80)_Fwd together with the respective gene binding reverse

348 primer. Escherichia coli TOP10 was used for molecular cloning. Primers used in this study are

349 listed in Supplementary Table S4.

350 Media and culturing conditions

351 Media. For E. coli culture, Luria Bertani medium containing 5 g/L yeast extract, 10 g/L

352 tryptone, and 5 g/L NaCl, 100 mg/L ampicillin as a selection was used. For Y. lipolytica

353 characterization, MpA defined media was used for all experiments. MpA components are as

354 follows: 5 g/L (NH4)SO4, 2 g/L KH2PO4, 0.5 g/L MgSO4, 44 mg/L ZnSO4·7H2O, 79 mg/L

355 CaCL2·2H2O, 0.8 mg/L Biotin, 100mM HEPES buffer, 100mM phosphate buffer (prepared as 0.9M

356 dibasic, 0.1M monobasic), trace elements (0.4 mg/L ZnSO4·7H2O, 0.04 CuSO4·5H2O, 0.4

357 MnSO4·6H2O, 0.2 mg/L Na2MoO4·2H2O, 0.1 mg/L KI, 0.5 mg/L FeSO4, 0.5 mg/L H3BO3), 380

358 mg/L leucine, 20 g/L glucose and various concentrations of thiamine hydrochloride. Initial media

359 pH was adjusted to 5pH.

360 Culturing. All experiments were conducted in a Kuhner LT-X incubator set to 28°C and

361 250 rpm unless otherwise stated. Fresh colonies were inoculated in 2mL of MpA medium

362 containing 400 µg/L thiamine in 15mL culture tubes overnight. Cultures were centrifuged and

363 resuspended in 2mL of water before transferring 1mL of this suspension into 100mL of MpA

364 containing 5 µg/L thiamine to scale-up cultures for 2 days (Figure 2A). Next, cells were washed

365 once with water and resuspended in 100mL of MpA lacking thiamine for 1 day to eliminate

366 thiamine carry-over (Figure 2A). Finally, cells were washed twice with water before

367 characterization experiments. All experiments were conducted in technical triplicates using

368 500mL baffled flasks unless otherwise stated.

369 Analytical methods

370 Real-time quantitative PCR (rt-PCR). Y. lipolytica, grown in MpA medium using glucose as

371 a carbon source together with either low (0.5 µg/L) and high (500 µg/L) thiamine, was collected

372 at mid-exponential phase (OD 2 ~ 3). Total RNA was purified by using the Qiagen RNeasy mini kit

373 (Cat # 74104, Qiagen Inc, CA, USA), and cDNA was subsequently synthesized by the QuantiTect

374 Reverse Transcription kit (Cat # 205311, Qiagen Inc, CA, USA). To quantify mRNA expression level

375 of genes (e.g., actin (YALI0D08272g), P1, P2, and P3), rt-PCR run was performed using the

376 QuantiTect SYBR Green PCR kit (Cat # 204143, Qiagen Inc, CA, USA) and StepOnePlus™ Real-Time

377 PCR System (Applied Biosystems, CA, USA). Primers used for rt-PCR are listed in Supplementary

378 Table S4. The gene expression level was investigated after normalizing by the actin house-keeping

379 gene as described elsewhere (23).

380 Promoter characterization with hrGFP. Fresh colonies of Y. lipolytica promoter

381 constructs were grown in 2 mL of MpA medium containing 400 µg/L thiamine overnight. Cultures

382 were washed once with water and transferred into 25mL of MpA containing 5 µg/L thiamine

383 overnight. Finally, cells were washed twice with water before being inoculated in MpA medium

384 with various concentrations of thiamine. Incubation was performed at 400 rpm and 28° using

385 96-well plates and Duetz-system covers (Cat# SMCR1296, Kuhner, Switzerland). Sacrificial

386 samples were collected for fluorescence measurements (excitation at 485 nm and emission at

387 528 nm) using a synergy HT microplate reader.

388 High performance liquid chromatography. Prior to HPLC run, 1 mL of culture medium

389 was filtered using 0.2 m filters. Metabolites, substrates and products were quantified by a

390 Shimadzu HPLC system equipped with UV and RID detectors (Shimadzu Scientific Instruments,

391 Inc., MD, USA) and the Aminex 87H column (Biorad, CA, USA) with 10 mN H2SO4 mobile phase at

392 0.6 mL/min flow rate. The column was maintained at 48 °C (23).

393 Proteomic analysis. Y. lipolytica were grown in biological triplicate in 0 and 400 µg/L

394 thiamine. Samples were collected at two time points during the exponential growth phase and

395 processed for LC-MS/MS analysis. Whole-cell lysates were prepared by bead beating in sodium

396 deoxycholate lysis buffer (4% SDC, 100 mM ammonium bicarbonate, pH 8.0) using 0.15 mM

397 zirconium oxide beads and cell debris cleared by centrifugation (21,000 x g for 10 min). Crude

398 protein concentrations were measured by a NanoDrop OneC (ThermoScientific) using

399 absorbance at 205 nm. Samples were then adjusted to 10 mM dithiothreitol and incubated at

400 85 °C for 10 min to denature and reduce proteins. Cysteines were alkylated/blocked with 30 mM

401 iodoacetamide followed by 20 min incubation at room temperature in the dark. Proteins (300 µg)

402 were then transferred to a 10-kDa MWCO spin filter (Vivaspin 500, Sartorius) and digested in situ

403 with proteomics-grade trypsin (Pierce) as previously described (57). The tryptic peptide solution

404 was then filtered through the MWCO membrane by centrifugation (12,000 x g for 15 min),

405 adjusted to 1% formic acid to precipitate SDC, and SDC precipitate removed from the peptide

406 solution with water-saturated ethyl acetate. Peptide samples were then concentrated to dryness

407 via SpeedVac, resolubilized in solvent A (5% acetonitrile, 95% water, 0.1% formic acid), and

408 measured by NanoDrop OneC A205 to assess tryptic peptide recovery.

409 Peptide samples were analyzed by automated 1D LC-MS/MS analysis using a Vanquish

410 UHPLC plumbed directly in-line with a Q Exactive Plus mass spectrometer (Thermo Scientific)

411 outfitted with a trapping column coupled to an in-house pulled nanospray emitter. Both the

412 trapping column (100 µm ID) and nanospray emitter (75 µm ID) were packed with 5 µm Kinetex

413 C18 RP resin (Phenomenex) to 10 cm and 30 cm, respectively. For each sample, 3 µg of peptides

414 were loaded, desalted, separated and analyzed across a 210 min organic gradient with the

415 following parameters: sample injection followed by 100% solvent A chase from 0-30 min (load

416 and desalt), linear gradient from 0% to 25% solvent B (70% acetonitrile, 30% water, 0.1% formic

417 acid) from 30-240 min (separation), followed by a ramp to 75% solvent B from 240-250 min

418 (wash), re-equilibration to 100% solvent A from 250-260 min and a hold at 100% solvent A from

419 260-280 min. Eluting peptides were measured and sequenced by data-dependent acquisition on

420 the Q Exactive MS as previously described (57).

421 MS/MS spectra were searched against the Y. lipolytica proteome concatenated with

422 common protein contaminants using Proteome Discover v.2.2 (ThermoScientific) employing the

423 CharmeRT workflow (58, 59). Peptide spectrum matches (PSM) were required to be fully tryptic

424 with 2 miscleavages; a static modification of 57.0214 Da on cysteine (carbamidomethylated) and

425 a dynamic modification of 15.9949 Da on methionine (oxidized) residues. False-discovery rates,

426 as assessed by matches to decoy sequences, were initially controlled at < 1% at both the PSM-

427 and peptide-levels. FDR-controlled peptides were then quantified by chromatographic area-

428 under-the-curve (AUC), mapped to their respective proteins, and areas summed to estimate

429 protein-level abundance. Protein abundance distributions were then normalized across samples

430 using InfernoRDN (60) and missing values imputed to simulate the MS instrument's limit of

431 detection using Perseus (61). Significant differences in protein abundance were calculated

432 separately for each time point according to the following equation:

WT_00t1,2 − WT_400t1,2 433 Fold change = Variance √0.25 + ∑ ⁄n

434 Here, WT_00 and WT_400 represent log2 normalized abundance of a protein in 0 and 400 µg/L

435 thiamine, respectively. The denominator was used to account for error between replicates.

436 Variance represents the variance of protein abundance between replicates, n represents the

437 number of replicates, and 0.25 is the pseudo variance term (62). Proteins with fold changes >

438 |1| were classified as upregulated or downregulated. Pathway annotations were performed

439 with ClueGo (63) using KEGG (https://www.genome.jp/kegg/) database on proteins that were

440 upregulated or downregulated at both time points.

441 Bioinformatics. Putative native Y. lipolytica thiamine-regulated promoters were

442 identified by BlastP (64) and orthologs search through the KEGG sequence similarity database

443 (https://www.kegg.jp/kegg/ssdb/). Reference genes used in this study were (i) P. pastoris

444 hydroxymethylpyrimidine phosphate synthase thi11 (PAS_chr4_0065) (65), (ii) S. pombe 4-

445 amino-5-hydroxymethyl-2-methylpyrimidine phosphate synthase nmt1 (NP_588347.1) (55), iii)

446 S. pombe thiamine thiazole synthase nmt2 (NP_596642.1) (66), and iv) S. cerevisiae thiamine

447 thiazole synthase thi4 (NP_011660.1) (67).

448 Lipid quantification. Thiamine-auxotrophic and thiamine-prototrophic strains were

449 cultured in 0 and 400 ug/L thiamine in MpA media in triplicates as outlined previously. Lipid

450 samples were taken from 100 uL samples of culture broth and incubated at room temperature

451 for 15 minutes in the dark after addition of 2 uL of 1 ug/mL BODIPY (cat # D3922, Fisher Scientific)

452 (68). Lipid standards were created by dissolving 100 mg of corn oil in 20mL ethanol and diluted

453 from 1-.1 mg/mL prior to BODIPY staining procedure. Lipids were measured using fluorescence

454 (ex: 485 nm/em: 528 nm) and quantified from corn oil standards. For dry cell weight (DCW)

455 measurement, 1mL of culture broth was sampled from each replicate at each time point.

456 Samples were centrifuged at maximum speed for 3 minutes and supernatant was discarded prior

457 to drying samples at 55°C overnight. DCW was calculated by subtracting the dried cell pellet and

458 tube weight by the empty tube weight. Finally, % lipid accumulation was calculated by dividing

459 the measured lipid mg/mL by DCW mg/mL. Statistical significance was calculated using SigmaPlot

460 14 with one-way analysis of variance (ANOVA) with Holm-Sidak correction between 0 and 400

461 µg/L thiamine for each strain respectively. Symbols: “*”: p-value < 0.05.

462

463 ACKNOWLEDGEMENTS

464 We would like to acknowledge financial support from the DOE BER Genomic Science

465 Program (DE-SC0019412) and the National Science Foundation (NSF #1511881). The views,

466 opinions, and/or findings contained in this article are those of the authors and should not be

467 interpreted as representing the official views or policies, either expressed or implied, of the

468 funding agencies.

469

470 FIGURE LEGENDS

471 Figure 1. Metabolic map of thiamine-dependent enzymes (green) in relationship to central

472 (black) and peripheral (blue) pathways. Abbreviations: TPP: thiamine pyrophosphate dependent

473 enzymes, PDH: pyruvate dehydrogenase enzyme complex, KGDH: a-ketoglutarate

474 dehydrogenase, TKL: transketolase, AHAS: acetolactate synthase, BCKDC: branched chain α-

475 ketoacid dehydrogenase, TPP: thiamine pyrophosphate, AAs: amino acids, TCA: citrate (Krebs)

476 cycle, KGA: alpha-ketoglutarate.

477 Figure 2. Growth characterization of the thiamine-auxotrophic Y. lipolytica YlSR001 in 0 (red),

478 0.5 (blue) and 400 (purple) µg/L thiamine. (A) Scheme of thiamine depletion design experiments.

479 (B) Cell growth profiles. (C) Glucose consumption profiles. (D) KGA production profiles. (E)

480 Pyruvate production profiles.

481 Figure 3. Proteomics of the thiamine-auxotrophic Y. lipolytica YLSR001 grown in 0 (red) and 400

482 (purple) µg/L thiamine. (A) Growth profiles and proteomics samples indicated by arrows. (B)

483 Venn diagram representing upregulated and downregulated proteins at both exponential time

484 points. (C) Thiamine-responsive proteomes on metabolic pathways. Bar graphs represent the

485 percentage of proteins upregulated in 0 (red) or 400 (purple) µg/L thiamine for each pathway (D)

486 Electron transport chain and ATPase. (E) Thiamine metabolism of Y. lipolytica. Locus tags: Thi4P:

487 YALI0A09768p, Thi20p: YALI0E04224p, Thi6p: YALI0C15554p, PHO: YALI0A12573p, Thi80p:

488 YALI0E21351p, ADK: YALI0F26521p). Abbreviations: PDH: pyruvate dehydrogenase enzyme

489 complex, KGDH: a-ketoglutarate dehydrogenase, TKL: transketolase, AHAS: acetolactate

490 synthase, BCKDC: branched chain α-ketoacid dehydrogenase, TPP: Thiamine pyrophosphate,

491 HET-P: hydroxyethylthiazole phosphate, HMP-PP: hydroxymethylpyrimidine pyrophosphate, TP:

492 thiamine monophosphate, TPP: thiamine pyrophosphate, TPPP: thiamine triphosphate.

493 Figure 4. Thiamine-responsive promoter characterization. (A) Chromosomal gene expression of

494 thiamine-responsive genes using rt-PCR. (B) hrGFP protein expression. (C, D) Sensitivity of

495 thiamine-responsive promoters in presence of increasing thiamine concentrations.

496 Figure 5. Expression of pP3-scTHI13 in YlSR1006 restores the de novo thiamine biosynthesis in Y.

497 lipolytica. (A) Cell growth profiles. (B) Glucose consumption profiles. (C) KGA production profiles.

498 (D) Pyruvate production profiles. Growth characterization was conducted in 0 (red), 0.5 (blue)

499 and 400 (purple) µg/L thiamine. Abbreviations: KGA, alpha-ketoglutarate.

500 Figure 6. Lipid production is influenced by thiamine availability. (A) Lipid accumulation profiles

501 for the thiamine-auxotrophic wildtype YlSR001 and thiamine-prototrophic strain YlSR1006 in 0

502 and 400 µg/L thiamine. (B) Lipid accumulation profiles for thiamine-auxotrophic and

503 prototrophic strains in lipid production (C:N = 100) media with 0 and 400 µg/L thiamine.

504 Statistical significance was calculated with one-way analysis of variance (ANOVA) with Holm-

505 Sidak correction between 0 and 400 µg/L thiamine for each strain respectively. Symbols: “*”: p-

506 value < 0.05.

507 Figure 7. A comprehensive model explains detrimental consequences of thiamine deficiency on

508 metabolism and cellular growth of Y. lipolytica.

509 Figure 1

510

511

512

513

514

515 Figure 2

516

A 0 µg/L 0.5 µg/L 400 µg/L Inoculum Scale-up Deplete Experiment 400 µg/L 5 µg/L 0 µg/L thiamine B C 12 20

9 no thiamine 15 0.5 g/L thiamine

600nm 6 400 g/L thiamine 10 OD

3 (g/L)Glucose 5

0 0 0 24 48 72 96 0 24 48 72 96 Time (h) Time (h) D E 2.0 5

4 1.5 3 1.0

2 KGA (g/L) KGA

0.5 (g/L)Pyruvate 1

0.0 0 0 24 48 72 96 0 24 48 72 96 517 Time (h) Time (h)

518

519 Figure 3

520

A C

400 µg/L thiamine 0 µg/L thiamine

B E

NADH Cytochrome c Cytochrome c dehydrogenase (I) reductase (III) oxidase (IV) ATPase ATPase ATP synthase (F-type) (V-type) (F-type) Succinate ATPase (V-type) dehydrogenase (II) Downregulated Cytochrome c oxidase (IV) Upregulated Cytochrome c reductase (III) X Succinate dehydrogenase (II) NADH dehydrogenase (I) 0 1 2 3 4 5 521 Number of proteins

522

523 Figure 4

524

A B 10 5.E+03 0.5 µg/L 1 µg/L 500 µg/L 10 mg/L

8 4.E+03 hrGFP 6 3.E+03

4 2.E+03

2 1.E+03 OD normalized OD normalized

Normalized expression gene Normalized 0 0.E+00 Actin P1 P2 P3 TEF NMT1 P1 P2 P3 Promoters Promoters C D

1.E+04 0.5 µg/L 12 0.5 µg/L 1 µg/L 1 µg/L 8.E+03 2 µg/L 10 2 µg/L 5 µg/L 5 µg/L 10 µg/L 8 10 µg/L 6.E+03 50 µg/L 50 µg/L

600nm 6

4.E+03 OD 4 2.E+03

OD normalized hrGFP hrGFP OD normalized 2

0.E+00 0 0 12 24 36 48 0 12 24 36 48 60 Time (h) Time (h) 525

526

527 Figure 5

528

A B 12 20

9 15

600nm 6 10

no thiamine OD

0.5 g/L thiamine Glucose (g/L)Glucose 3 400 g/L thiamine 5

0 0 0 24 48 72 96 0 24 48 72 96 Time (h) Time (h) C D 2.5 1.2

2.0 0.9

1.5 0.6

1.0

KGA (g/L) KGA Pyruvate (g/L)Pyruvate 0.3 0.5

0.0 0.0 0 24 48 72 96 0 24 48 72 96 529 Time (h) Time (h)

530

531

532 Figure 6

533

A B 10 No nitrogen limitation 10 Nitrogen limitation

24h 24h 8 8 48h 48h 72h 72h 6 6

4 4 * * *

* % lipid accumulation % lipid 2 accumulation % lipid 2 * * * * * * * * 0 0 400WT g/L 0 WTg/L 400P3 g/L 0 P3g/L 400WT g/L 0 WTg/L 400P3 g/L 0 P3g/L 400 µg/LWildtype0 µg/L 400 µg/LYlSR10060 µg/L 400 µg/LWildtype0 µg/L 400 µg/LYlSR10060 µg/L 534

535

536

537

538 Figure 7

539

Consequences of Thiamine Deficiency DNA, RNA, NADPH Antioxidants, Lipids, Ribosome, RNA Purines, Pyrimidines and DNA deficiency Oxidative Stress Ribose-5P Reduced carbon assimilation Xylulose-5P Glycolysis TKL X X X Aromatic AAs Erythrose-4P Folate

Seduheptulose-7P TDL Auxotroph

X TKL

X AHAS X Pyruvate Auxotroph X Branched PDH Acid production chain AAs X X BCKDC X Acetyl-CoA Lipids, Sterols

Deficient lipid NADH production Energy (ATP) Succinyl-CoA TCA KGA deficiency

Electron Transport KGDH Chain Acid production 540

541

542 REFERENCES

543 1. Brown GM (1970) Biogenesis and Metabolism of Thiamine. in Metabolic Pathways, eds

544 Greenberg DM, Vogel HJ, & Hopkin LE (New York, Academic Press, New York), pp 369-

545 378.

546 2. Vemuganti R, Kalluri H, Yi J-H, Bowen KK, & Hazell AS (2006) Gene expression changes in

547 thalamus and inferior colliculus associated with inflammation, cellular stress, metabolism

548 and structural damage in thiamine deficiency. European Journal of Neuroscience

549 23(5):1172-1188.

550 3. Bunik VI, Tylicki A, & Lukashev NV (2013) Thiamin diphosphate-dependent enzymes: from

551 enzymology to metabolic regulation, drug design and disease models. The FEBS Journal

552 280(24):6412-6442.

553 4. Linn TC, Pettit FH, & Reed LJ (1969) α-KETO ACID DEHYDROGENASE COMPLEXES, X.

554 REGULATION OF THE ACTIVITY OF THE PYRUVATE DEHYDROGENASE COMPLEX FROM

555 BEEF KIDNEY MITOCHONDRIA BY PHOSPHORYLATION AND DEPHOSPHORYLATION.

556 Proceedings of the National Academy of Sciences 62(1):234-241.

557 5. Patel MS, Nemeria NS, Furey W, & Jordan F (2014) The pyruvate dehydrogenase

558 complexes: structure-based function and regulation. J Biol Chem 289(24):16615-16623.

559 6. Gibson GE, Blass JP, Beal MF, & Bunik V (2005) The α-ketoglutarate-dehydrogenase

560 complex. Molecular Neurobiology 31(1):43-63.

561 7. Reed LJ (1974) Multienzyme complexes. Accounts of Chemical Research 7(2):40-46.

562 8. Schenk G, Duggleby RG, & Nixon PF (1998) Properties and functions of the thiamin

563 diphosphate dependent enzyme transketolase. The International Journal of Biochemistry

564 & Cell Biology 30(12):1297-1318.

565 9. Boros LG, et al. (2003) Defective RNA ribose synthesis in fibroblasts from patients with

566 thiamine-responsive megaloblastic anemia (TRMA). Blood 102(10):3556-3561.

567 10. Rapala-Kozik M, Kowalska E, & Ostrowska K (2008) Modulation of thiamine metabolism

568 in Zea mays seedlings under conditions of abiotic stress. Journal of Experimental Botany

569 59(15):4133-4143.

570 11. Pettit FH, Yeaman SJ, & Reed LJ (1978) Purification and characterization of branched chain

571 α-keto acid dehydrogenase complex of bovine kidney. Proceedings of the National

572 Academy of Sciences 75(10):4881-4885.

573 12. Ciszak EM, Korotchkina LG, Dominiak PM, Sidhu S, & Patel MS (2003) Structural Basis for

574 Flip-Flop Action of Thiamin Pyrophosphate-dependent Enzymes Revealed by Human

575 Pyruvate Dehydrogenase. Journal of Biological Chemistry 278(23):21240-21246.

576 13. WILLIAMS RD, MASON HL, SMITH BF, & WILDER RM (1942) INDUCED THIAMINE (VITAMIN

577 B1) DEFICIENCY AND THE THIAMINE REQUIREMENT OF MAN: FURTHER OBSERVATIONS.

578 JAMA Internal Medicine 69(5):721-738.

579 14. Organization WH (1999) Thiamine deficiency and its prevention and control in major

580 emergencies. (Department of Nutrition for Health and Development, Geneva,

581 Switzerland).

582 15. Liu D, Ke Z, & Luo J (2017) Thiamine Deficiency and Neurodegeneration: the Interplay

583 Among Oxidative Stress, Endoplasmic Reticulum Stress, and Autophagy. Molecular

584 Neurobiology 54(7):5440-5448.

585 16. Bettendorff L, et al. (1995) Thiamine Deficiency in Cultured Neuroblastoma Cells: Effect

586 on Mitochondrial Function and Peripheral Benzodiazepine Receptors. Journal of

587 Neurochemistry 64(5):2013-2021.

588 17. Whitfield KC, et al. (2018) Thiamine deficiency disorders: diagnosis, prevalence, and a

589 roadmap for global control programs. Annals of the New York Academy of Sciences

590 1430(1):3-43.

591 18. Du Q, Wang H, & Xie J (2011) Thiamin (vitamin B1) biosynthesis and regulation: a rich

592 source of antimicrobial drug targets? Int J Biol Sci 7(1):41-52.

593 19. Groenewald M, et al. (2014) Yarrowia lipolytica: Safety assessment of an oleaginous yeast

594 with a great industrial potential. Critical Reviews in Microbiology 40(3):187-206.

595 20. Markham KA & Alper HS (2018) Synthetic Biology Expands the Industrial Potential of

596 Yarrowia lipolytica. Trends in Biotechnology 36(10):1085-1095.

597 21. Hussain MS, et al. (2016) Recent advances in bioengineering of the oleaginous yeast

598 Yarrowia lipolytica. Aims Bioeng 3:493-514.

599 22. Ryu S, Hipp J, & Trinh CT (2016) Activating and elucidating metabolism of complex sugars

600 in Yarrowia lipolytica. Appl. Environ. Microbiol. 82(4):1334-1345.

601 23. Ryu S & Trinh CT (2018) Understanding functional roles of native pentose-specific

602 transporters for activating dormant pentose metabolism in Yarrowia lipolytica. Appl.

603 Environ. Microbiol. 84(3):e02146-02117.

604 24. Schwartz C, Curtis N, Löbs A-K, & Wheeldon I (2018) Multiplexed CRISPR Activation of

605 Cryptic Sugar Metabolism Enables Yarrowia Lipolytica Growth on Cellobiose.

606 Biotechnology Journal 13(9):1700584.

607 25. Walker C, Ryu S, & Trinh CT (2019) Exceptional solvent tolerance in Yarrowia lipolytica is

608 enhanced by sterols. Metabolic Engineering 54:83-95.

609 26. Ryu S, Labbé N, & Trinh CT (2015) Simultaneous saccharification and fermentation of

610 cellulose in ionic liquid for efficient production of α-ketoglutaric acid by Yarrowia

611 lipolytica. Applied Microbiology and Biotechnology 99(10):4237-4244.

612 27. Xu P, Qiao K, & Stephanopoulos G (2017) Engineering oxidative stress defense pathways

613 to build a robust lipid production platform in Yarrowia lipolytica. Biotechnology and

614 Bioengineering 114(7):1521-1530.

615 28. Zhou J, Zhou H, Du G, Liu L, & Chen J (2010) Screening of a thiamine-auxotrophic yeast for

616 α-ketoglutaric acid overproduction. Letters in Applied Microbiology 51(3):264-271.

617 29. Morgunov IG, Kamzolova SV, Perevoznikova OA, Shishkanova NV, & Finogenova TV (2004)

618 Pyruvic acid production by a thiamine auxotroph of Yarrowia lipolytica. Process

619 Biochemistry 39(11):1469-1474.

620 30. Li M, et al. (2010) Thiamine Biosynthesis in Saccharomyces cerevisiae Is Regulated by the

621 NAD+-Dependent Histone Deacetylase Hst1. Molecular and Cellular Biology 30(13):3329-

622 3341.

623 31. Hohmann S & Meacock PA (1998) Thiamin metabolism and thiamin diphosphate-

624 dependent enzymes in the yeast Saccharomyces cerevisiae: genetic regulation.

625 Biochimica et Biophysica Acta (BBA) - Protein Structure and Molecular Enzymology

626 1385(2):201-219.

627 32. Landes N, et al. (2016) The vitamin-sensitive promoter PTHI11 enables pre-defined

628 autonomous induction of recombinant protein production in Pichia pastoris.

629 Biotechnology and Bioengineering 113(12):2633-2643.

630 33. Belén Moreno M, Durán A, & Carlos Ribas J (2000) A family of multifunctional thiamine-

631 repressible expression vectors for fission yeast. Yeast 16(9):861-872.

632 34. Basi G, Schmid E, & Maundrell K (1993) TATA box mutations in the Schizosaccharomyces

633 pombe nmt1 promoter affect transcription efficiency but not the transcription start point

634 or thiamine repressibility. Gene 123(1):131-136.

635 35. Müller S, Sandal T, Kamp-Hansen P, & Dalbøge H (1998) Comparison of expression

636 systems in the yeasts Saccharomyces cerevisiae, Hansenula polymorpha, Klyveromyces

637 lactis, Schizosaccharomyces pombe and Yarrowia lipolytica. Cloning of two novel

638 promoters from Yarrowia lipolytica. Yeast 14(14):1267-1283.

639 36. Damude HG, et al. (2011) High eicosapentaenoic acid producing strains of Yarrowia

640 lipolytica. (Google Patents).

641 37. Scott M, Gunderson CW, Mateescu EM, Zhang Z, & Hwa T (2010) Interdependence of Cell

642 Growth and Gene Expression: Origins and Consequences. Science 330(6007):1099-1102.

643 38. Klumpp S, Zhang Z, & Hwa T (2009) Growth Rate-Dependent Global Effects on Gene

644 Expression in Bacteria. Cell 139(7):1366-1375.

645 39. Rodríguez-Navarro S, et al. (2002) Functional analysis of yeast gene families involved in

646 metabolism of vitamins B1 and B6. Yeast 19(14):1261-1276.

647 40. Maundrell K (1993) Thiamine-repressible expression vectors pREP and pRIP for fission

648 yeast. Gene 123(1):127-130.

649 41. Lonsdale D (2006) A Review of the Biochemistry, Metabolism and Clinical Benefits of

650 Thiamin(e) and Its Derivatives. Evidence-Based Complementary and Alternative Medicine

651 3(1):49-59.

652 42. Wolak N, Kowalska E, Kozik A, & Rapala-Kozik M (2014) Thiamine increases the resistance

653 of baker's yeast Saccharomyces cerevisiae against oxidative, osmotic and thermal stress,

654 through mechanisms partly independent of thiamine diphosphate-bound enzymes. FEMS

655 Yeast Research 14(8):1249-1262.

656 43. Singleton CK & Martin PR (2001) Molecular Mechanisms of Thiamine Utilization. Current

657 Molecular Medicine 1(2):197-207.

658 44. Dimmer KS, et al. (2002) Genetic Basis of Mitochondrial Function and Morphology in

659 Saccharomyces cerevisiae. Molecular Biology of the Cell 13(3):847-853.

660 45. Fattal-Valevski A (2011) Thiamine (Vitamin B1). Journal of Evidence-Based

661 Complementary & Alternative Medicine 16(1):12-20.

662 46. Sundström M, Lindqvist Y, Schneider G, Hellman U, & Ronne H (1993) Yeast TKL1 gene

663 encodes a transketolase that is required for efficient glycolysis and biosynthesis of

664 aromatic amino acids. Journal of Biological Chemistry 268(32):24346-24352.

665 47. Kispal G, Steiner H, Court DA, Rolinski B, & Lill R (1996) Mitochondrial and Cytosolic

666 Branched-chain Amino Acid Transaminases from Yeast, Homologs of the myc Oncogene-

667 regulated Eca39 Protein. Journal of Biological Chemistry 271(40):24458-24464.

668 48. Shabbir Hussain M, Gambill L, Smith S, & Blenner MA (2016) Engineering Promoter

669 Architecture in Oleaginous Yeast Yarrowia lipolytica. ACS Synthetic Biology 5(3):213-223.

670 49. Trassaert M, et al. (2017) New inducible promoter for gene expression and synthetic

671 biology in Yarrowia lipolytica. Microbial Cell Factories 16(1):141.

672 50. Blazeck J, Liu L, Redden H, & Alper H (2011) Tuning Gene Expression in Yarrowia lipolytica

673 by a Hybrid Promoter Approach. Applied and Environmental Microbiology 77(22):7905-

674 7914.

675 51. Nicaud J-M, et al. (2002) Protein expression and secretion in the yeast Yarrowia lipolytica.

676 FEMS Yeast Research 2(3):371-379.

677 52. Tikka C, et al. (2013) Isolation and characterization of ethanol tolerant yeast strains.

678 Bioinformation 9(8):421-425.

679 53. Gibson DG, et al. (2009) Enzymatic assembly of DNA molecules up to several hundred

680 kilobases. Nat Meth 6(5):343-345.

681 54. Yamane T, Sakai H, Nagahama K, Ogawa T, & Matsuoka M (2008) Dissection of

682 centromeric DNA from yeast Yarrowia lipolytica and identification of protein-binding site

683 required for plasmid transmission. Journal of Bioscience and Bioengineering 105(6):571-

684 578.

685 55. Moreno MB, Duran A, & Ribas JC (2000) A family of multifunctional thiamine-repressible

686 expression vectors for fission yeast. Yeast (Chichester, England) 16(9):861-872.

687 56. Wang JH, Hung WP, & Tsai SH (2011) High efficiency transformation by electroporation of

688 Yarrowia lipolytica. J Microbiol 49(3):469-472.

689 57. Clarkson SM, et al. (2017) Construction and Optimization of a Heterologous Pathway for

690 Protocatechuate Catabolism in Escherichia coli Enables Bioconversion of Model Aromatic

691 Compounds. Applied and Environmental Microbiology 83(18):e01313-01317.

692 58. Villalobos Solis MI, Giannone RJ, Hettich RL, & Abraham PE (2019) Exploiting the Dynamic

693 Relationship between Peptide Separation Quality and Peptide Coisolation in a Multiple-

694 Peptide Matches-per-Spectrum Approach Offers a Strategy To Optimize Bottom-Up

695 Proteomics Throughput and Depth. Analytical Chemistry 91(11):7273-7279.

696 59. Dorfer V, Maltsev S, Winkler S, & Mechtler K (2018) CharmeRT: Boosting Peptide

697 Identifications by Chimeric Spectra Identification and Retention Time Prediction. Journal

698 of Proteome Research 17(8):2581-2589.

699 60. Taverner T, et al. (2012) DanteR: an extensible R-based tool for quantitative analysis of -

700 omics data. Bioinformatics 28(18):2404-2406.

701 61. Tyanova S, et al. (2016) The Perseus computational platform for comprehensive analysis

702 of (prote)omics data. Nature Methods 13:731.

703 62. Mukhopadhyay A, et al. (2006) Salt Stress in Desulfovibrio vulgaris: an Integrated

704 Genomics Approach. Journal of Bacteriology 188(11):4068-4078.

705 63. Bindea G, et al. (2009) ClueGO: a Cytoscape plug-in to decipher functionally grouped gene

706 ontology and pathway annotation networks. Bioinformatics (Oxford, England)

707 25(8):1091-1093.

708 64. Altschul SF, et al. (1997) Gapped BLAST and PSI-BLAST: a new generation of protein

709 database search programs. Nucleic acids research 25(17):3389-3402.

710 65. Delic M, Mattanovich D, & Gasser B (2013) Repressible promoters – A novel tool to

711 generate conditional mutants in Pichia pastoris. Microbial Cell Factories 12(1):6.

712 66. Manetti AG, Rosetto M, & Maundrell KG (1994) nmt2 of fission yeast: a second thiamine-

713 repressible gene co-ordinately regulated with nmt1. Yeast (Chichester, England)

714 10(8):1075-1082.

715 67. Praekelt UM, Byrne KL, & Meacock PA (1994) Regulation of THI4 (MOL1), a thiamine-

716 biosynthetic gene of Saccharomyces cerevisiae. Yeast (Chichester, England) 10(4):481-

717 490.

718 68. Klapper M, et al. (2011) Fluorescence-based fixative and vital staining of lipid droplets in

719 Caenorhabditis elegans reveal fat stores using microscopy and flow cytometry

720 approaches. Journal of Lipid Research 52(6):1281-1293.

721